Paraoxonase 2 (PON2) Deficiency Reproduces Lipid Alterations of Diabetic and Inflammatory Glomerular Disease and Affects TRPC6 Signaling

Abstract

Diabetes and inflammatory diseases are associated with an altered cellular lipid composition due to lipid peroxidation. The pathogenic potential of these lipid alterations in glomerular kidney diseases remains largely obscure as suitable cell culture and animal models are lacking. In glomerular disease, a loss of terminally differentiated glomerular epithelial cells called podocytes refers to irreversible damage. Podocytes are characterized by a complex ramified cellular architecture and highly active transmembrane signaling. Alterations in lipid composition in states of disease have been described in podocytes but the pathophysiologic mechanisms mediating podocyte damage are unclear. In this study, we employ a genetic deletion of the anti-oxidative, lipid-modifying paraoxonase 2 enzyme (PON2) as a model to study altered cellular lipid composition and its effects on cellular signaling in glomerular disease. PON2 deficiency reproduces features of an altered lipid composition of glomerular disease, characterized by an increase in ceramides and cholesterol. PON2 knockout mice are more susceptible to glomerular damage in models of aggravated oxidative stress such as adriamycin-induced nephropathy. Voltage clamp experiments in cultured podocytes reveal a largely increased TRPC6 conductance after a membrane stretch in PON2 deficiency. Correspondingly, a concomitant knockout of TRPC6 and PON2 partially rescues the aggravated glomerular phenotype of a PON2 knockout in the adriamycin model. This study establishes PON2 deficiency as a model to investigate the pathophysiologic mechanisms of podocyte dysfunction related to alterations in the lipid composition, as seen in diabetic and inflammatory glomerular disease. Expanding the knowledge on these routes and options of intervention could lead to novel treatment strategies for glomerular disease.

Keywords: calcium signaling; lipid peroxidation; oxidative stress; podocyte.

Figure 1

PON2 expression is induced in diabetic and inflammatory human glomerular kidney disease. Human kidney biopsy samples stained with PON2 specific antibody. PON2 reactivity is primarily localized in tubular cells in healthy kidney as well as in hypertensive (HTN) and minimal change (MCGN) glomerular disease while glomeruli show faint staining. PON2 staining is pronounced in glomeruli in diabetic kidney disease (DKD), ANCA vasculitis (ANCA), and lupus nephritis (SLE). Specifically, PON2 expression is predominant in podocytes (arrow heads) localizing to the outer aspect of the glomerular tuft.

Figure 2

PON2 deficiency alters lipid composition of cultured podocytes. Cellular sphingo- and phospholipids content of PON2-deficient and -proficient murine podocytes was assessed by LC MS/MS. PON2-deficient podocytes (grey bars) contain more ceramide and sphingomyelin compared to the control (black bars) (ceramide: mean 0.183 pmol/µg protein; SD 0.037 vs. mean 0.076 pmol/µg protein; SD 0.042; * p < 0.05; sphingomyelin: mean 0.581 pmol/µg protein; SD 0.059 vs. mean 0.240 pmol/µg protein; SD 0.035; ** p < 0.01). Content of phosphatidylcholine, phosphatidylethanolamine, and phosphatidylinositol are markedly reduced in PON2-deficient podocytes (* p < 0.05; ** p < 0.01; # not significant).

Figure 3

PON2 deficiency alters lipid composition in adriamycin-induced nephropathy kidney cortex samples of adriamycin-treated mice were subjected to LC MS/MS to quantify ceramide content and subclasses. (A) Total ceramide content is significantly increased in PON2 ko (triangles) as compared to wildtype mice (dots) (n = 8 and 5, respectively). (B) Both C16/C18 ceramide and C22/C24 ceramide are increased in PON2 ko compared to control. (C) Expression profiling of PON2-proficient and -deficient podocytes investigating RNA levels of NPHS1, ceramide generating enzymes Cers6 and SMPD1/2, and ceramidase enzymes Asah1 and Acer2. (D) Acid sphingomyelinase (ASM) activity was assessed in cortex samples of PON2 ko and wildtype mice. ASM activity is largely increased in PON2 ko vs. PON2 wt (n = 6 and 5, respectively) (** p < 0.01; *** p < 0.001).

Figure 4

PON2 deficiency aggravates the glomerular phenotype of adriamycin-induced nephropathy. Quantification of albuminuria in the AIN model on day 0, day 14, and day 49 after adriamycin induction (A). No albuminuria is detected at baseline. Wildtype (circles, n7) and PON2 heterozygous mice (squares, n6) develop subnephrotic range albuminuria on day 14 and recover to baseline on day 49. PON2 ko mice (triangles, d14 n9, and d49 n6) show nephrotic albuminuria on day 14 which aggravates until day 49. Two-way ANOVA with p < 0.001 for genotype, days after treatment and interaction. (* p < 0.05; ** p < 0.01) (B,C) Serum creatinine and serum urea on day 14 and 49. At day 49 serum creatinine is significantly increased in PON2 ko mice compared to wildtype and PON2 heterozygous mice (* p < 0.05; n = 5 animals each group). No significant differences for serum urea levels are detected. (D) Representative images of histologic analysis of kidney sections on day 49 do not reveal glomerular sclerosis in any of the samples. In PAS stainings, glomerular adhesion to the Bowman’s capsule (arrowhead) and foam cell degeneration of podocytes (arrow) are present in samples of PON2 ko animals. (E) Representative electron microscopy images of kidney samples on day 49 show foot process effacement on PON2 ko mice whereas wildtype mice present regular foot process architecture.

Figure 5

TRPC6 conductance is regulated by PON2. (A) Immunoprecipitation of FLAG-tagged PON2 and V5-tagged slit diaphragm proteins reveals interaction of PON2 and TRPC6, Podocin, and Nephrin and no interaction with CD2AP. (B) Representative recordings of whole cell voltage clamp experiments using PON2-proficient and -deficient murine podocytes. TRPC6 currents in response to membrane stretch after hypotonic stimulation (70%) are largely increased in PON2-deficient podocytes. (C) Currents at +80 mV during whole cell recordings including medium change from 100% tonicity to 70% tonicity and back. In PON2-deficient podocytes currents stay active for 40 min, while control cells recover after 5 min. (D) Enhanced TRPC6 conductance of PON2-deficient podocytes is reduced by re-expression of PON2 wildtype (* p < 0.05) (control mean 0.39 nA, SD 0.16; PON2 shRNA mean 4.26 nA, SD 0.73; PON2 shRNA + PON2 wt 2.73 nA, SD 0.26).

Figure 6

Aggravation of the adriamycin-induced nephropathy phenotype in PON2 knockout is mediated by TRPC6. (A) Quantitative analysis of albuminuria in PON2/TRPC6 double knockout (squares) and PON2 ko/TRPC6 heterozygous (circles) mice. While PON2ko/TRPC6het mice develop albuminuria on day 14, PON2ko/TRPC6ko mice are protected (n = 4 and 5, respectively; * p < 0.05, # not significant). In the further course, mice of both genotypes show progressive nephrotic proteinuria. (B) Quantification of serum creatinine on day 49 show increased concentrations in PON2ko/TRPC6het mice as compared to PON2ko/TRPC6ko (n = 3; * p < 0.05).